Sunday, September 17, 2023

When predictions go wrong - the defensive responses

 


Predictions will go wrong. The likelihood that any forecaster will be better than a flip of a coin is rare, yet most predictors will come up with some reason for their failure.  It is not me; it is the market that got it wrong. Here are some of the key defenses for why a forecaster will say it is not him that is the problem. 

1. The "if only" defense - If only the Fed did what I said they should have done, I would have been right. Those players are not being rational or at least not the way I think they should be acting.

2. The "ceteris paribus" defense - All things equal, I would have been correct, but there was an event that was not included in the model. Those unforeseen events caused the problem. It was not my model.

3. The "I was almost right" - what I predicted did not occur, but I was close enough. Yes, I was wrong, but it was close to what was expected. It was within a range of tolerance albeit it was still wrong.

4. The " It just hasn't happened yet" defense - I am correct, but it is taking longer than expected. You can get the direction or timing right but not both. Take your pick. Most will not get both right.

5. The "single prediction" defense - Ok, I was wrong, but this is just an isolated event. Mistakes were made. We are only human, so I will get it right the next time.

-paraphrased from James Montier


No comments: